‘This is the process power part of the equation,’ Larry said. ‘It goes like this: the first step is to define
what the problems are…’
‘Where the shoe pinches’, I interjected.
‘Exactly ‘, he said. ‘But the key is to focus not simply on the outcome, which almost everyone does, but
on the process that creates the product or service. Then, map that process so that you can easily recognize
the links between the steps! Sometimes that’s where the problems are – not in the teeth, but in the gaps
between the teeth.’
‘So the goal is not only to brush, but also to floss.’
‘Hey not bad!’ Larry said. ‘Mind if I steal that one?’
‘Be my guest’, I offered magnanimously.
‘Another key : don’t just complain about things that can’t be fixed, like the weather. Focus on problems
that you can fix. There's no point cursing the darkness, as they say. It’s not a bitch session. The key is to
define the problems objectively…’
‘Bingo! If we say, “Our quality is inconsistent”, or “Customers don’t like the New-York-style pizza”, what
does that mean? How do you make quality more consistent or fix the darn New-York-style pizza with data
like that? The more accurately you define the problem, the more precise your target, the better your
chances are for hitting the bull’s-eye.’
‘When you gather data on the problem, you have to do it carefully, because you're writing the instruction
booklet for the rest of the project. Do a sloppy job here, and your chances for success are as slim as your
chances for programming your VCR correctly the first time using those awful instruction books.’
‘If you go to the doctor and just say, “I don't feel well”, and he doesn’t ask any more questions or conduct
any tests, he’s probably not going to get too far in trying to make you feel better. Clarity, clarity, clarity!
And it starts with clearly identifying the problem.’
‘Think about it’, he continued, clearly on a roll. ‘When effective armies go to the war, they
don’t say, “See if you can drop a bomb somewhere in enemy territory”. They say, “Attack
the military base, not the bakery, and it’s located at these coordinates”. Then, you know
exactly where to strike!’
‘And you know, it’s interesting. Sometimes just soliciting the problems from your customers and employees
and making them feel better – because someone’s finally listening to them – and the act of writing the
problems down makes them feel more manageable. Instead of an ill-defined grey cloud hanging over your
company, your problems become clear, specific items you can identify and correct.’
‘It’s reassuring to see that problems can be broken down, listed and attacked. Instead of just looking down at
your belly and saying. “ I’m really out of shape”, you look at a cold, clear number indicating how much you
weigh and write down another number indicating where you want to go. Suddenly, it doesn’t seem so
‘So, you list all the problems you’ve got …’ I said, getting him back to the point.
‘As many as you think are necessary’, Larry corrected. ‘Start with what the customers think, then go on to
what the guys think who deal directly with the customer or with the machines that create the product.’
‘Okay’, I said. ‘Then what?’
‘Pick the problem that’s giving you the most trouble, the one that’s costing the company
the most, the one that’s making customers unhappy – the one that will reward you the
most if you can fix it’, Larry said. ‘We’re not looking for carpet-bombing here but a surgical
strike. Bombing the bakery might deprive the Citizens of a few jam doughnuts, but it won’t
get at the core of the problem. You need to attack the shipyard.’
‘In our business, if we want to get the pizza to the customer faster, we don’t bother trying to speed up
our delivery cars’, he explained. ‘Sure, they might save an extra second or two on acceleration when the
light turns green, but for the money it would cost us, it would hardly be worth it. If we focus instead on
speeding up the cooking process, the we’re on to amore cost-effective solution.’
‘Makes sense’, I said. ‘So, you’ve got the problem.’
‘Then you go on to the second step’, he said. ‘You measure.’
‘Measure what?’ I asked.
‘Lots of things’, Larry said. ‘Let’s go back to the doctor’s office for a minute. When you say you don’t feel
good, a good doctor won’t just say, “Okay, let’s run these tests”, because she’d have no idea to which tests
to run and would waste a lot of time and money going down blind alleys.‘
Instead, she’ll ask you where it hurts, when it hurts, and a whole lot of follow-up questions, so she can begin
to narrow down the possibilities. Only then does she start running tests.
‘Likewise,’ Larry continued, ‘once you’ve got the right tests set up, it’s time to measure
the capability of a given process – what’s possible – by measuring how many
opportunities for defects a certain process or operation presents.’
In baseball terms, it’d be equivalent to keeping track of how many chances for errors a fielder has – how
many fly balls or grounders come their way. From there the Black Belt calculates how many errors are made,
which is called the frequency of defects.
‘Put the second number over the first,’ Larry explained, ‘and you've got your fielding percentage.
Most companies measure the number of errors, but not the number of opportunities, so they
don’t know what's possible, and how far they’re falling short.’
‘Next, do some benchmarking by measuring the competition’s fielding percentage on that problem’, he said.
‘How are they doing on the same problem you’ve picked? How does that compare to your company’s
performance? Before most companies start breaking it down by numbers, they usually assume they’re one
of the best in the field of quality, efficiency and customer satisfaction. And why not? They’ve usually been
in the business for a while – and for a good reason. But after they look around and compare themselves
with what the competition’s doing, they usually discover that they’re not as exceptional as they thought.’
Lawrence Bossidy, the former Allied Signal CEO I mentioned earlier, tells his staff they should assume
that everyone of Allied’s competitors does at least one thing better than Allied. After all, they must be doing
something right, or they wouldn’t be in the business.
‘The concept’s pretty simple. If you’re pro golfer and you want to improve, you break down your game
piece by piece. How well do you drive, how well do you chip, how well do you putt? Then, compare your
numbers to the best driver, the best chipper, the best putter. How do you measure up? And most importantly,
what is the best golfer in each area doing that you’re not? It doesn’t matter if their overall score is better than
yours, what matters is if they can teach you something in one part of the game.’
‘Let’s say our problem is consistence’, Larry said. ‘Let’s find out who has the best consistency in the
business – or even observe companies in another field if that helps. Benchmarking tells us what’s possible
and gives us a reasonable goal. Then, we workout what they’re doing that we’re not.’
‘During the entire measurement process,’ he added, ‘it’s important for the Black Belt to focus on the
“critical to quality”, or CTQ, characteristics – those that have the most impact on the outcome. For example,
A team’s players may not be very tall compared to another team’s, but that fact may not be critical quality.
How they field ground balls, however, surely will be. So, the Black Belt needs to find out which elements
are critical to quality.’
I was no Black Belt, of course, but it wasn’t hard to understand that figuring out
which elements were critical to quality – in this case, catching and throwing the
ball – was vital to improving your fielding.
‘Of all steps to Six Sigma,’ Larry said, ‘the measurement phase is probably the most underestimated,
in terms of its importance and the time and money it takes to do it right. Because it’s not that flashy step,
And nothing really happens in measurement, there’s a tendency to zip through it without much thought.’
‘But that’s a big mistake, because good numerical data is the foundation of Six Sigma. Without good data,
you can’t make good decisions. There’s and old carpenter’s maxim: If you think it will take one day and
$100, plan for two day and $200. The same is true for measurement.’
‘I’m with you’, I said. ‘So you’ve got all this data. What’s next?’
‘If the data’s good,’ he said, ‘you can start the next step: analyze the numbers to find out how
well or poorly the processes are working, compared with what’s possible and with what the
competition’s doing. Done right, the first three steps will show you the maximum results
possible if everything is perfect, and also how far your company is falling short.’
If the gap between the two isn’t great, you don’t have much to gain from improving your performance, so move
on to next problem. If the gap is great, start digging in. You're on to something valuable.
The big question to answer, of course, are why the errors are big committed and how to fix them. Is it poor
technique fielding fly balls, bad throwing arms, or not enough range to field ground balls? If you set the
Experiment upright, the numbers will tell you the answers.
‘If you don’t get the numbers you need, go back to the drawing board and set up a new experiment – the
same way a doctor would order more tests.' Larry mentioned.
‘Sounds great,’ I said., ‘but what are you looking for? How do you know when you’ve struck gold?’
‘If you can answer when, where, and how often the defects occur, you have what you need’, Larry said.
‘But don’t just focus on the symptoms of the problem. Find the root causes. If you stop looking for the
problem halfway, you’ll come up with a half – baked solution to eradicate it.’
‘Let’s say we play by the rules’, I postulate. ‘We identify a juicy problem, we measure how we are doing,
and analyze the numbers. What do we do with them?’
‘You decide how far off the mark the numbers are’, Larry answered. ‘Once you know what you really
weigh, you decide what your target weight is and put together a plan of attack to get there, including
a diet, exercise programme and deadline.’
‘But a business just ain’t that simple’, I said.
‘True enough’, Larry replied. ‘That’s why the Six Sigma classes teach statistical reasoning, among
other topics, to help the Black Belts compile and use the data in the most effective way possible.
Responding to that data is the next step: improve the processes you’re pursuing.’
‘What’s the message?’ I asked. ‘Are we back to, “Hey, everyone, could you please do your jobs better?”’’
‘No, it’s more sophisticated than that – thank God!’ Larry laughed. ‘Since we’ve already identified which
CTQ components are not meeting our expectations, we then focus on implementing changes that will
improve the specific CTQs. This will improve the whole process.’
Let’s say we’ve worked out, for example, that fielding is essential to winning base ball games, and that
handling ground balls is essential to fielding well – and further, that we could do a lot better at that phase of
the game. Then, it stands to reason that if we improve our ability to field ground balls, our fielding percentage
will go up, and our winning percentage will follow.
‘In business terms, you sit down with the data and you analysis and determine what’s possible
– what’s a reasonable goal for this or that process. After you’ve established numerical standards, you
calculate which of your “tolerance bands” need to be tightened to get there and then how to go about
In the case of base ball team, let’s say we’re winning an average of 80 games a year, but we’ve
determined that with our personnel and facilities and budget, we should be able to win 90. To do so, we’ll
need to bring our fielding average up from 85% to 95%. And, to do that, we’ll need to field 90 out of
100 ground balls successfully, not just the 75 we currently are fielding without errors.
Now , how do we do that? Bring in a fielding coach to work on the techniques we lack. That coach – the
Black Belt – works out specifically what we can be doing better to field more ground balls cleanly –
nuts-and-bolts things like bending our knees and keeping our gloves down on the ground.’
‘It’s a long way from “Please try harder”’, I said.
‘Glad you agree’, Larry said.
‘So is that it?’ I asked.
‘Not quite’, he answered. ‘It’s a bit like losing weight…’
‘Back to weight loss, eh?’
‘Just for a moment. A lot of programmes help a lot of people take off the weight, but the hard part is
keeping it off.’
‘Don’t I know it!’ I said, patting my paunch.
‘Exactly’, Larry said, patting his in a show of brother-hood. ‘The last step is control, in
which you lock in your successes.’
‘And how do you do that?’
The Black Belts implement measures to keep the key variables within their new
operating limits month after month’, Larry said. ‘In this step, though, it’s important to
distinguish between statistical process monitoring and statistical process control.
It’s the difference between hopping on the scales every morning to check your weight
and watching your calories every day to control your weight.’
‘I know that difference all too well’, I admitted.
‘Don’t we all’, he said with a resigned grin. ‘Once the plan’s in place, it’s the Master Black Belts job to
make sure the team monitors the process, measures the results and confirms that the plan is working. He
or she also has to monitor variables that might be affecting the data, things they never thought about
on the drawing board. If any new problems arise, they refine the process and reload.’
‘To cut a long story short,’ Larry said, ‘you Define the problem, Measure where you stand, Analyze where
the problem starts, Improve the situation and Control the new process to confirm that it’s fixed – DMAIC, or
Dumb Managers Always Ignore Customers.’
‘I feel like I’m playing the piano again’, I said.
‘Sorry’, Larry said. ‘But after my training, I had to score 80% or better on the final test to be
certified as a Black Belt. You do everything you can to remember the information.’
‘Hey, don’t tell me where you got the matches’, I said. ‘Just tell me how big the fire was!’
Larry had a good laugh at his own expense. ‘Point well taken, my friend’, he said. ‘But the upshot is
pretty simple. We measure what we care about. If we didn’t value home runs, we wouldn’t count them.
Do we count foul balls? We start to measure it, then we begin to improve our performance on it.’
‘Second, stick a date to everything you do. Without deadlines, it just doesn’t get done. If the tax office said,
‘Hey, just get your taxes in when it’s convenient for you, what do you think would happen?’’
And third, assign very specific jobs to everyone on your team. That eliminates the confusion and lack
of direction we talked about earlier, and also guarantees that the job gets done. Even the Red Cross
advices rescuers on the scene to point to a specific member in the crowd to get a blanket, to another to
call the emergency services, and so on. Otherwise, everyone just stands there.
‘Finally, put it all in writing and pass it around. If everyone can see who’s doing what and when, that
breeds some accountability.’
“Pick the problem that’s giving you the most trouble, the one that’s costing the company the most,
the one that’s making customers unhappy – the one that will reward you the most if you can fix it.”
“The big questions to answer are why the errors are being committed and how to fix them.”
“Sit down with the data and your analysis and determine what’s possible – what’s a reasonable
goal for this or that process.”
“Define the problem, Measure where you stand, Analyze where the problem starts, improve
the situation and Control the new process to confirm that it’s fixed – DMAIC, or Dumb Managers
Always Ignore Customers.”
Now that Joe Meter is convinced at last with the benefits Six Sigma can bring to any
place, lets see if Larry can get his complete buy-in with a real life application…
Let’s see it happen (as we are now sure Larry/Six Sigma can do it) next week.
Thanks for your time